Do not modify the input astropy table when converting to pandas dataframe

[jasongrout]

How could the content be improved?

In #74, the function that makes a pandas dataframe from an astropy table was updated to modify the input astropy table (adding some columns). In general, it's probably not a good idea to modify an input to a function unless the function is specifically designed to modify its input in place. However, I understand the change was made to resolve some errors in a pandas version update.

It seems that using .value when adding a skycoord sequence to pandas works now, e.g., df['phi1'] = skycoord_gd1.phi1.value:

def make_dataframe(table):
    """Transform and astropy table with coords in ICRS, convert to pandas dataframe with GD-1 coordinates."""
    skycoord = SkyCoord(
        ra=table['ra'],
        dec=table['dec'],
        pm_ra_cosdec=table['pmra'],
        pm_dec=table['pmdec'],
        distance=8*u.kpc,
        radial_velocity=0*u.km/u.s)

    # Transform to GD-1 frame
    gd1_frame = GD1Koposov10()
    transformed = skycoord.transform_to(gd1_frame)
    skycoord_gd1 = reflex_correct(transformed)

    df = table.to_pandas()
    df['phi1'] = skycoord_gd1.phi1.value
    df['phi2'] = skycoord_gd1.phi2.value
    df['pm_phi1'] = skycoord_gd1.pm_phi1_cosphi2.value
    df['pm_phi2'] = skycoord_gd1.pm_phi2.value

    return df

Which part of the content does your suggestion apply to?

https://datacarpentry.github.io/astronomy-python/03-transform.html#pandas-dataframe

[jasongrout]

CC @abostroem, who recommended I open an issue about this. Thanks!